Learning Treatment Algorithms from Generative AI and Unstructured Medical Data

07/2024 – 06/2025 | SNU Convergence Research Program

• Development of a generative AI-based algorithm to refine and standardize unstructured data, which accounts for 80% of all medical data
• Construction of a medical-specialized large language model (LLM) trained on clinicians’ treatment algorithms to enable clinical reasoning tasks such as summarization, diagnosis support, and treatment prediction
• Validation of practical applicability and productivity improvements in real clinical settings in collaboration with medical professionals at Seoul National University Hospital